Prostate cancer is the second leading cause of cancer deaths among men in the United States, and the vast majority of these deaths are the result of androgen-independent disease. Currently there are no effective means to prevent the development of androgen-independent prostate cancer or to treat patients once they have developed this deadly disease. This proposal aims to: 1) identify differentially expressed proteins associated with androgen-independent prostate cancer and, 2) validate clinical significance of differential expression of these proteins by examining a large series of human surgical specimens. The CWR22 xenografi and human surgical specimens will be used in this study. Laser capture microdissection will be used to procure pure populations of specific cancerous cells. Protein expression patterns will be compared by 2D-PAGE and the identity of differentially expressed proteins will be determined by mass spectrometry sequencing. The frequency of altered protein expression and the relationship to cellular proliferation and clinical evidence of progression to androgen independence will be assessed by immunohistochemical studies of the complete CWR22 xenograft series as well as serial biopsies obtained from advanced prostate cancer patients prior to and at intervals following castration. The long-term goal of this proposal is to find new biomarkers that can be used to identify men with androgen-independent prostate cancer clones prior to the onset of clinical androgen independence. This will facilitate development of therapies that prevent progression to androgen-independent disease and will enable clinicians to identify those men who would benefit from these strategies. For those men who already have androgen-independent clones, prevention strategies will be ineffective, and novel therapies directed at killing or controlling androgen-independent clones will need to be developed. Protein changes discovered by this proposal will likely be good targets for these cytotoxic or suppressive therapies.